Projects

GUIDANCE conducts research on general-purpose, dialogue-assisted digital information access — enabling users to access digital information through dialogue while overcoming key limitations of current LLMs (they were not designed for information access, generalise poorly to new domains and languages, and their output veracity is hard to guarantee). The project addresses four challenges: designing or re-using LLMs for dialogue-based information access; leveraging retrieval-enhanced machine learning to improve retrieval accuracy and efficiency; adapting LLMs to low-resource settings and languages (e.g. Occitan, French); and ensuring the veracity and explainability of retrieved and synthesised information while preserving the user's subjectivity. GUIDANCE is led by Benjamin Piwowarski (CNRS, ISIR/MLIA) with IRIT, LIG and LIS, and aims to federate the French Information Retrieval community.

Project website · ANR record

In the CoST project, we envision a shift from search engines to task completion engines by dynamically assisting users in making the optimal decisions empowering them to achieve multi-step and highly cognitive search tasks. This triggers the need for (1) more predictable and automatic models of user-system interactions and search tasks and, (2) more task-oriented information access models. The objectives envisioned in the CoST project are: (1) Identifying patterns of users' behaviours while completing complex search tasks. Our aim here is to discover behavioural regularities across users and relate them through clustering techniques that could explain the nature of the involved task; (2) learning explicit and structured representations of complex search tasks, based on those behavioural patterns. Our objective here is to capture the relationships and the dependencies between task stages, i.e., the overall structure of tasks; (3) modelling task-driven IR by relating document relevance to task completion. The driving idea here is to leverage from the search patterns on the one hand and the structure of tasks on the other hand, to establish possible user actions and rewards corresponding to the accomplishment of the task. The main scientific rupture intended in the CoST project is the definition of theoretical foundations of task-based IR, a radically new IR approach

More details can be found on the project website